Pharmacokinetic and biodistribution analysis of monoclonal antibodies: a comprehensive study of antibody biodistribution and partitioning coefficients in mice

ABSTRACT

The efficacy of therapeutic monoclonal antibodies (mAbs) often hinges on biodistribution to their site of action. However, traditional pharmacokinetic (PK) assessments – typically based on measuring plasma or total tissue concentrations – fail to reflect the interstitial concentrations that are most relevant for tissue targets. This study aimed to address this limitation by integrating experimentally determined vascular and interstitial volumes from tissues in SCID-beige mice with a comprehensive PK time-course and biodistribution analysis of four distinct anti-viral monoclonal antibodies (mAbs 1–4) with no endogenous mouse target. The biodistribution studies included 11 tissues, characterizing tissue and plasma concentrations over a 168-h time-course. Total and interstitial tissue concentrations were evaluated to better understand concentrations within the interstitial space compared to bulk tissue values. These data revealed significant tissue-specific partitioning, with fold-change analysis suggesting groupings correlating with capillary endothelium characteristics. A dynamic model was implemented for the estimation of antibody biodistribution coefficient (ABC) values at steady-state, partitioning ratio (PR) values at steady-state, and their associated equilibrium rate constants (t_1/2eq, t’_1/2eq) across 11 (ABC, t_1/2eq) and 7 tissues (PR, t’_1/2eq), respectively. Specifically, to understand non-binding, target-independent biodistribution, we combined data from mAbs 1, 2, and 3 to create a “typical mAb” (mAb 123) profile, from which these coefficients and ratios were derived. Analysis of mAb 4, a structurally similar IgG molecule with undesirable PK properties, enabled comparative insights into antibody distribution and kinetics. These studies provided a comprehensive dataset for understanding interstitial antibody PK, crucial for improving predictions of PK at the site-of-action and in vivo efficacy.

Introduction

The development of therapeutic antibodies has revolutionized modern medicine by offering targeted treatments for various diseases ranging from viral infections to autoimmune disorders. However, their clinical success hinges not only on systemic exposure but also on achieving sufficient concentrations at the relevant site of action. Traditional preclinical studies often fall short in capturing tissue-level distribution. By improving our understanding of inherent tissue partitioning in animal models, we can better predict therapeutic efficacy and optimize dosing. Pharmacokinetic (PK) properties, encompassing absorption, distribution, metabolism, and excretion (ADME), are critical determinants of therapeutic efficacy and safety of new drugs. Therapeutic antibodies exert their main effects in the interstitial space of tissues, a characteristic stemming from their drug properties. Therefore, their efficacy is directly tied to their biodistribution into this interstitial compartment.¹ From a PK perspective, the distribution of antibodies is characterized by the volume of distribution (V_d), which is influenced by the volume of plasma and tissue (V_p, V_t) and the tissue-to-plasma partitioning (Kp).² Following absorption into the plasma compartment, antibodies move into tissue interstitial space through diffusion or convection depending on the type of capillary endothelium (continuous, fenestrated, sinusoid, or tight junctions).³

Measuring antibody PK within the interstitium is generally not feasible due to the challenge of experimentally sampling the sub-compartment level of tissues. Instead, typical preclinical PK studies rely on traditional methods of measuring plasma or total tissue homogenate concentrations, which fail to capture the effective concentration within the interstitial space. Total tissue homogenates includes the sum of antibody levels resulting from plasma, interstitial fluid, and cells within a tissue sample, which underestimates the exposure to the interstitial space.⁴ Tissue concentrations are often extrapolated from plasma or serum values using physiologically based pharmacokinetic (PBPK) models or using an antibody biodistribution coefficient (ABC) approach for understanding tissue distribution.^5–8 PBPK models rely on known physiological and chemical parameters to extrapolate tissue concentration from plasma data whereas ABC values provide a simple method for estimate of total tissue concentration from plasma concentration based on ratios derived from numerous reported literature values.⁹ Although straightforward and easy to implement, the accuracy and precision of these approaches are constrained. This limitation stems from the fact that the literature values underpinning ABC calculations often represent just a single snapshot in time, thereby failing to capture the full dynamic PK profile of the therapeutic within tissues.

Radiolabeled antibodies, often incorporating gamma-emitting isotopes such as Iodine-125, offer distinct advantages for studying antibody PK and biodistribution in vivo due to their precise quantification capabilities.¹⁰ From the values obtained from gamma counting, one can readily derive the plasma, total tissue, and interstitial concentration of an antibody over time in various tissues if certain physiology parameters are known (i.e., the vascular and interstitial volume).^11,12 In contrast to extrapolated methods, simultaneously measuring plasma and total tissue concentrations and then calculating the interstitial concentration offers a highly accurate means of assessing tissue distribution. This detailed approach yields valuable insights into the complex relationship between an antibody’s PK and its pharmacodynamic effects.

Previous studies characterizing antibody biodistribution have often incorporated reported values from other studies, leading to confounding results due to a lack of standardization across the studies.^8,13 The variation in studies used to create biodistribution values and models include mouse and rat models of different age and genetic backgrounds, different radioisotope probes and antibodies that have distinct distribution due to wide-ranging properties.⁸ Factors that affect antibody tissue PK parameters include antibody-specific physiochemical and functional properties (charge, hydrophobicity, target affinity, FcRn affinity, Fc receptor interactions, and glycosylation), target properties (expression level, turnover rate, and soluble versus membrane-associated), drug administration (dose and route), anti-therapeutic antibody formation, off-target/nonspecific binding, and disease state.^14,15 Given the many factors that could affect the PK parameters of therapeutic antibodies, finding and characterizing a preclinical species in which to assess the biodistribution of antibodies at the tissue and sub-compartmental levels in the body is critical. Further variation comes from distinctive sampling methods, assorted bioanalytic techniques with varied sensitivity, and propagation of literature values in their derivation to assess biodistribution. While the variety in experiments may help capture the general trend of biodistribution, it fails to capture a precise assessment of antibody biodistribution.

To address the conflicting reports in the literature,^8,13 a primary aim of our study was to generate data that elucidates the plasma to interstitial fluid concentration partitioning ratio (PR) of mAbs in the liver. Furthermore, we extended our in vivo measurements of IgG antibody concentrations to the interstitial compartment of other tissues. Alongside our investigation of liver partitioning, we collected and analyzed a broader panel of tissues and organs to generate a comprehensive dataset of IgG distribution over time.

We chose to characterize our molecules biodistribution in a commonly used mouse model for assessing preclinical PK, the severe combined immunodeficiency (SCID) beige mouse.¹⁶ This mouse model is often used to avoid immunogenicity caused by administering human or humanized antibodies to preclinical species, which affects PK interpretation. We carried out our PK studies using radiolabeled antibodies with a non-residualizing radioisotope probe, and we chose four internal antiviral antibodies to assess the biodistribution of an antibody without influence of target present. These antibodies are against four different viruses: mAb1 - respiratory syncytial virus (RSV), mAb2 - hepatitis B virus (HBV), mAb3 - herpes simplex virus (HSV), and mAb4 - human immunodeficiency virus (HIV). All four molecules are human IgG1 isotype. Three antibodies (mAb 1, 2 and 3) are classic full-length IgG molecules with typical well-behaved PK profiles in mouse. One antibody, mAb 4, has abnormal PK and distribution due to its glycosylation profile leading to faster clearance. Specifically, mAb 4 exhibits lower levels of sialic acid on its Fab glycan (data not shown), presumably making it a substrate for the liver asialoglycoprotein receptor (ASGPR).¹ For this reason, we selected mAb 4 as a comparative tool molecule to understand distribution and kinetics in contrast to mAbs 1–3. Three of these antibodies (mAbs 1–3) retain their native Fc regions, enabling effector functions such as antibody-dependent cellular cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC).¹⁷ The fourth antibody includes the DLE and LS modifications in its Fc region, which are designed to enhance Fc-mediated effector function.^18,19

Here, we calculated interstitial concentrations using measured physiological values in combination with PK data from the radiolabeled antibody study. In addition to improving tissue-level characterization, we also sought to evaluate the reliability of traditional bioanalytical methods for quantifying total tissue concentrations. Specifically, we compared ligand-binding assay (LBA) measurements with radiolabeled biodistribution data to assess whether LBA can accurately reflect tissue exposure. This comparison helps illustrate the limitations of conventional approaches and supports the use of radiolabeling for more comprehensive biodistribution analysis. To our knowledge, this is the most comprehensive study in terms of the number of tissues and organs characterized in combination with PK aimed at measuring interstitial concentrations over time. This study generates comprehensive antibody biodistribution coefficients (ABC, total tissue concentration to serum concentration ratios) values and tissue partitioning ratios (PR, calculated as interstitial tissue concentration to serum concentration ratios) across multiple time points, offering a dynamic view of antibody distribution. The overall workflow and experiments are shown in Scheme 1. These data can serve as refined inputs to PBPK models, enhancing their translational accuracy by better capturing the temporal and spatial aspects of tissue exposure. Rather than serving as alternatives, these measurements complement and refine existing data by providing more physiologically relevant details on tissue partitioning and interstitial exposure. Hence, our results can help reduce uncertainty in estimating site-of-action concentrations.

Scheme 1.

Overview of the workflow for characterizing vascular and interstitial volumes (Vv, Vi) and conducting radiolabeled biodistribution studies using monoclonal antibodies (mAbs 1–4). Experimental data were used to calculate interstitial mAb concentrations in tissues over time, as well as ABC (ratio of total tissue to serum concentration) and PR (ratio of interstitial tissue to serum concentration). Steady-state ABC, PR, and equilibrium half-lives (Keq) were derived from these data. Abbreviations: Cp, central compartment; Ct, peripheral compartment; SoA, site of action; V, volume of site of action; Cl, clearance; Q, intercompartmental distribution; Vv, vascular volume; Vi, interstitial volume; Vc, cellular volume.

Materials and methods

All animal experiments, including both the radiolabeled and non-radiolabeled PK studies, were conducted in accordance with applicable ethical guidelines and regulatory requirements. The radiolabeled PK study was performed under a license authorized by the National Animal Experiment Board of Finland and followed the National Institutes of Health guidelines for the care and use of laboratory animals. The non-radiolabeled PK study was conducted under IACUC-approved guidelines at LabCorp Madison, ensuring compliance with institutional and national standards for animal care and use.

Non-radiolabeled PK analysis

Female SCID-beige mice (CRL, Germany), age 6–8 weeks of age, were dosed at 1 mg/kg via an intravenous (IV) bolus injection through the tail vein. Individual doses were calculated based on body weights recorded on the day of administration. Prior to dosing, the mAb stock solutions were diluted in 1 ×phosphate-buffered saline (PBS), pH 7.4 to achieve a dose volume of 2 mL/kg. Sparse sampling was used to optimize blood collection from individual mice, minimizing animal stress and ethical considerations while enabling comprehensive PK parameter estimation. This approach involved collecting a limited number of blood samples from each mouse at predefined, staggered time points, with the complete PK profile subsequently reconstructed by integrating data obtained from multiple animals within the same experimental group. Blood was collected from three mice per group per time point via submandibular puncture or cardiac puncture. Blood was allowed to clot at ambient temperature prior to centrifugation to obtain serum. Samples were centrifuged at 2,200 ×g within 1 h of collection and frozen at −80°C until analysis. The concentration of all mAbs were measured using a generic Meso Scale Discovery (MSD) electrochemiluminescence (ECL) based immunoassay detection method of sufficient selectivity and sensitivity. Briefly, mAbs in serum were captured using biotinylated anti-human IgG (Southern Biotech; Cat. No.: 2049–08) coated on streptavidin GOLD 96-well plates and detected using SULFO-TAG^TM conjugated anti-human IgG detection antibody (MSD). After incubations and washings, a read buffer is added, and plates are analyzed on an MSD QuickPlex SQ 120 plate reader. The ECL signal produced is proportional to the amount of mAb bound to the plate. The concentrations of mAb in control serums and samples are back calculated with a standard curve using a 4-parameter logistic curve fit.

Measurement of tissue vascular volume

The vascular volumes of mouse tissues were measured by technetium-99 m (^99mTc) labeling of red blood cells (RBCs) in vivo using the indirect labeling method following the administration of stannous (Sn²⁺) pyrophosphate a component of the reconstituted TechneScan PYP kit, as described by Boswell et al.²⁰ The TechneScan PYP kit was dissolved in 6 ml saline (4 mg of tin chloride in 6 ml of normal sterile saline). A 75 µL aliquot of this solution (corresponding to 50 µg of tin chloride) was diluted with 10 mL saline. A 100 µL dose was drawn from this solution and injected (IV) via tail vein to a donor mouse. This was followed exactly 30 min later by an IV bolus dose of ^99mTc pertechnetate. One hour after the radioactive dose, the donor mouse received deep isoflurane anesthesia, and blood was collected using cardiac puncture. Collected blood of the donor mouse was transferred to an ethylenediaminetetraacetic acid (EDTA)-coated tube, and immediately 150 µl of ^99mTc-RBC were drawn and dosed to naïve recipient mice. A sample of donor blood was centrifuged at 2,000 ×g for 10 min at 4°C, and plasma and pellet radioactivity were measured to ensure non-RBC associated levels of ^99mTc.²¹ The radioactivity of the dosing syringe was measured prior and after dosing to obtain exact dosed activity. One hour later, the recipient animals were terminally anesthetized with pentobarbital (180 mg/kg). Next, blood was collected using cardiac puncture, and animals were decapitated followed by tissue collection. Collected tissues were rinsed with PBS to remove external blood and blotted before weighing and counted for radioactivity. Plasma samples were centrifuged at 2,000 ×g for 10 min at 4°C.

The radioactivity of collected samples was measured using a gamma counter. The radioactivity amount in the tissue, blood and plasma was decay corrected to report %ID/g. Vascular volume (Vv) was calculated using ^99mTc-RBC results using Equation 1.

$$\mathit{Vv}\left(\mathit{\mu }\mathrm{l}\mathrm{blood}/\mathrm{g}\mathrm{tissue}\right)=\frac{\mathit{CPM}/g\left(\mathit{tissue}\right)}{\mathit{CPM}/\mathit{\mu }l\left(\mathit{blood}\right)}$$ (1)

Measurement of tissue interstitial volume

The interstitial volumes of mouse tissues were measured by continuous infusion of the extracellular marker, indium-111 diethylenetriaminepentaacetic acid (¹¹¹In-DTPA), as described by Boswell et al.²² Both male and female SCID-beige mice, age 6–8 weeks of age, were used. ¹¹¹In -DTPA was purchased from Curium Finland (Helsinki, Finland). On the day of the dosing, ¹¹¹In-DTPA solution was diluted with saline to obtain a solution that contained 1 MBq ¹¹¹In-DTPA in 300 µL solution. Upon dilution, animals were anesthetized and cannulated, and ¹¹¹In-DTPA (1 MBq) was administered by 1 h IV infusion at a rate of 300 µl/h. Dosed radioactivity was verified by collecting blank infusion to the tube and measuring the collected activity. Upon completion of the infusion, blood was immediately collected (within one minute after the end of infusion) via cardiac puncture, weighed into EDTA coated tubes, and centrifuged at 2,000 ×g for 10 min at 4°C to obtain plasma samples. Animals were decapitated followed by tissue collection. Collected tissues were rinsed with PBS to remove external blood and blotted before weighing and counting for radioactivity.

Interstitial volume (Vi) was calculated using ¹¹¹In-DPTA results and Vv values from above. Calculation was performed using Equation 2, where MBq/µL (blood) was the collected blood of endpoint sample from the ¹¹¹In-DPTA study.

$$\mathit{Vi}\left(\mathit{\mu }\mathrm{l}\mathrm{plasma}/\mathrm{g}\mathrm{tissue}\right)\mathit{\hspace{0.17em}}=\frac{\mathit{CPM}/g\left(\mathit{tissue}\right)-CPM/\mathit{\mu }l\left(\mathit{blood}\right)\ast Vv\left(\mathit{\mu }l\mathrm{blood}/g\mathrm{tissue}\right)}{\mathit{CPM}/\mathit{\mu }l\left(\mathit{plasma}\right)}$$ (2)

Radiolabeling of antibodies with Iodine-125

For tissue distribution studies, mAbs 1–4 were radiolabeled with iodine-125 (¹²⁵I) using the Iodogen method. Radiolabeling was performed with Pierce pre-coated iodination tubes (Thermo Fisher Scientific), following the “Direct method for Iodination” in manufacturer’s instructions (MAN0016379 Rev. A.0, Thermo Scientific).²³ Briefly, 130 µL of 25 mM TRIS HCl − 0.4 M NaCl buffer (pH 7.5), 6 µL of antibody solutions and approximately 20 MBq of Iodine-125 was add to the iodination tube. The tubes were mixed and then left to react for 15 min with occasional mixing. The reaction was stopped by removing the mixture from the iodination tube. Antibody mixture was purified using Zeba™ Spin Desalting Column (Thermo Fisher) and sterile filtered 1xPBS. Radiochemical purity of purified ¹²⁵I-antibody solution was determined with instant thin layer chromatography using 85% MeOH as mobile phase. In this system, radiolabeled protein had Rf = 0. The solution volume was adjusted so one dose was approximately 1 mg/kg and 5 mL/kg. To characterize extracellular and interstitial antibody distribution, we used a non-residualizing radiolabel (¹²⁵ I), which is known to be cleared following intracellular catabolism. This labeling strategy was intentionally selected to avoid confounding intracellular retention and to better reflect antibody behavior in the interstitial space. As such, tissue signal may underestimate total uptake in organs with high catabolic activity, and this limitation was considered in our interpretation of tissue kinetics.

Tissue distribution study and PK analysis of radiolabeled antibodies

All animal experiments were performed as specified in the license authorized by the National Animal Experiment Board of Finland and according to the National Institutes of Health guidelines for the care and use of laboratory animals. Experiments were conducted using female SCID-beige mice aged 6–8 weeks. Animals were randomly assigned to the treatment groups using a stratified randomization scheme designed to achieve balanced groups based on bodyweight.

Prior to dosing, the animals were habituated to the restraining tubes for 3 days, 0.5–1 min per day. 2 MBq in 100 µL of ¹²⁵I-antibody solution was dosed to each animal by IV bolus. The dose volume for each animal was based on individual body weight. The terminal body weight of the animals was recorded, and the animals were terminally anesthetized with pentobarbital (180 mg/kg). Next, blood was collected using cardiac puncture, and animals were decapitated followed by tissue collection at each time point. Collected tissues were rinsed with PBS to remove external blood and blotted before weighing and counted for radioactivity. Plasma samples were collected following centrifugation at 2,000 ×g for 10 min at 4°C.

Dose normalized total blood, serum, and tissue concentration was calculated using Equation 3. The %ID/g represents the percentage of the injected dose per gram of tissue; $\mathit{CP}{M}_{\mathit{tissue}}$ and $\mathit{CP}{M}_{\mathit{ID}}$ being the radioactivity in counts per minute within the collected sample and the total injected dose; mass in grams of sample for tissue samples was obtained by weighing the sample; and for blood and plasma the mass was obtained by using the density of the biofluid multiplied by the volume.

$$\%\mathit{ID}/\mathit{g}=\frac{\mathit{CP}{\mathit{M}}_{\mathit{sample}}}{\mathit{CP}{M}_{\mathit{ID}}\ast \mathit{mass}\mathit{in}\mathit{grams}\mathit{of}\mathit{sample}}$$ (3)

Equation 4 was used to calculate the concentration of antibody in each sample. The %ID/g was converted to its decimal form and multiplied by the dose (in µg).

$$C_t\left(\mathit{\mu g}/\mathit{g}\right)=\%\mathit{ID}/\mathit{g}\ast 0.01\ast \mathit{dose}$$ (4)

To calculate the concentration of antibody in tissue resulting from blood within the sample, Equation 5 was used.

$$C_{{\mathit{t}}_{\mathit{blood}}}\left(\mathit{\mu g}/\mathit{g}\right)=C_{\mathit{blood}}\ast \mathit{Vv}$$ (5)

This was then subtracted from the total tissue concentration to obtain the concentration of antibody in tissue, corrected for blood contamination, as shown in Equation 6.

$$C_{\mathit{t}\mathit{blood}\mathit{corrected}}\left(\mathit{\mu g}/\mathit{g}\right)=C_t-C_{{\mathit{t}}_{\mathit{blood}}}$$ (6)

Finally, the concentration of antibody in the interstitial space was calculated using Equation 7, using the blood corrected tissue concentration and dividing it by the interstitial volume.

$$C_i\left(\mathit{\mu g}/\mathit{g}\right)==\frac{{\mathit{C}}_{\mathit{t}\mathit{blood}\mathit{corrected}}}{V_i}$$ (7)

An exponential limited-growth equation was used to estimate the time courses of the antibody biodistribution coefficient (ABC(t)) and the partitioning ratio (PR(t)):

$$\mathit{ABC}\left(\mathit{t}\right)=\mathit{ABCss}\left(1-\mathit{exp}\left(-\mathit{K}.\mathit{t}\right)\right)$$ (8)

$$\mathit{PR}\left(\mathit{t}\right)=\mathit{PRss}\left(1-\mathit{exp}\left(-K^{\prime }.\mathit{t}\right)\right)$$ (9)

Where ABCss and PRss represent the equilibrium biodistribution and partitioning ratios coefficients at quasi-steady-state, respectively, and K and K′ are the rate constants governing the speed at which equilibrium is reached. At t = 0, ABC(0) = 0 and PR(0) = 0, and as time progresses, they asymptotically approach their steady-state values. For each tissue, the parameters were estimated separately by fitting the model to the time course of the experimental data for each mAb group. In addition, parameter estimation was performed using a bootstrapping approach (1,000 iterations) to quantify uncertainty. More details are provided in the Supplemental Material. For simplicity, throughout this paper, these steady-state values (ABCss and PRss) will be referred to simply as ABC and PR. ABC and PR are reported as percentages (%).

The equilibrium half-life was calculated for both ABC and PR (t_1/2eq and t’_1/2eq) using the rate constants K and K’ using Equation 10.²⁴

$$t_{1/2eq}=\frac{ln\left(2\right)}{\mathit{K}\left(\mathit{or}\mathit{K}{\mathit{ }}^{\mathrm{\prime }}\right)}$$ (10)

The time unit throughout this analysis was h; therefore, K and K’ are expressed in units of 1/hr, and t_1/2 is in h.

Results

Vascular and interstitial volume

Two independent studies were conducted to assess the vascular (Vv) and interstitial (Vi) volumes across various tissues in both male and female mice. In the first study, female mice were used, and the following tissues and matrices were collected: bladder, bone, bone marrow, brain, heart, kidney, large intestine, small intestine, liver, lungs, plasma, blood, spleen, and thyroid. In the second study, male mice were used. In addition to the organs listed above, brown adipose tissue, eye, inguinal fat, lymph node, skeletal muscle, skin, and testes were collected.

The Vv was measured using the ^99mTc-RBC assay presented in Figure 1A and the results are shown in Figure 1B. Using this methodology, only the Vv measurements of the heart showed a significant difference between males and females, with males having a higher mean Vv (126 ± 32.4 μL blood/g tissue) compared to females (71.4 ± 21.1 μL blood/g tissue). All other tissues showed similar vascular volumes between sexes.

Figure 1.

Vascular and interstitial volume quantification across multiple mouse tissues. (A) Schematic representation of the vascular volume assay. (B) Vascular volume measurements (mean ± SD, µL blood/g tissue) in various tissues. (C) Schematic representation of the interstitial volume assay. (D) Interstitial volume measurements (mean ± SD, µL plasma/g tissue) in various tissues. Significant differences between groups are indicated (*p < 0.05, **p < 0.01) (n = 6/tissue/sex).

The Vi was measured using the ¹¹¹In-DTPA assay outlined in Figure 1C, with Vi results shown in Figure 1D. In these studies, Vi measurements showed more notable sex-specific differences than the Vv. For example, in the heart, lungs, and large intestine, females had a significantly higher Vi (196 ± 31.6, 209 ± 13.2, and 138 ± 36.4 plasma/g tissue, respectively) compared to males (120 ± 31.7, 190. ± 5.63, and 77.4 ± 11.0 μL plasma/g tissue, respectively). Similar values were observed between sexes for all other organs (Figure 1D). The bladder and kidney were excluded from Vi determination because the radioisotope probes accumulated in these organs through renal elimination, which yielded physiologically implausible values (Fig. S1A). Additionally, the Vi for the brain was excluded due to the limitation of ¹¹¹In-DTPA crossing the blood–brain barrier (Figure S1B).²⁵

A detailed summary of these values is shown in Table 1. While the Vv of the heart and the Vi of the heart, lungs, and large intestine exhibited significant sex-specific differences, there was no strong evidence of broad or uniform sexual dimorphism across all other assessed volumes in this study. These experiments yielded a highly comprehensive characterization of vascular and interstitial volumes, representing one of the most extensive datasets reported to date.

Table 1.

Summary values of vascular (Vv) and interstitial (Vi) volumes.

Tissue	Sex	Vv (Mean + SD)	Vi (Mean + SD)
Brown Adipose Tissue	M	36.8 ± 7.74	73.3 ± 6.08
Bladder	F	40.8 ± 12.6	NA
	M	42.4 ± 17.1	NA
Bone	F	11.0 ± 7.94	130 ± 41.5
	M	14.7 ± 2.52	120. ± 18.7
Bone Marrow	F	44.9 ± 14.3	74.7 ± 17.1
	M	30.8 ± 6.62	97.5 ± 15.7
Brain	F	8.86 ± 1.42	NA
	M	10.3 ± 0.65	NA
Eye	M	9.84 ± 4.81	244 ± 33.4
Heart	F	71.4 ± 21.1	196 ± 31.6
	M	126 ± 32.4	120 ± 31.7
Inguinal Fat	M	11.5 ± 2.53	111 ± 24.1
Kidney	F	134 ± 16.8	NA
	M	113 ± 14.9	NA
Large Intestine	F	6.94 ± 1.16	138 ± 36.4
	M	5.94 ± 2.35	77.4 ± 11.0
Liver	F	54.0 ± 9.15	70.5 ± 10.9
	M	56.0 ± 10.7	65.8 ± 7.41
Lungs	F	161 ± 31.2	209 ± 13.2
	M	174 ± 13.7	190. ± 5.63
Lymph Node	M	17.8 ± 7.54	258 ± 80.9
Muscle	M	105 ± 19.9	7.16 ± 0.12
Skin	M	6.70 ± 2.28	280 ± 40.3
Small Intestine	F	144 ± 34.2	14.0 ± 3.07
	M	122 ± 15.4	15.1 ± 3.59
Spleen	F	164 ± 23.6	10.0 ± 3.32
	M	146 ± 29.7	22.3 ± 10.3
Testes	M	7.81 ± 0.68	162 ± 71.1
Thyroid	F	47.2 ± 24.4	215 ± 97.2
	M	74.0 ± 18.4	215 ± 87.6

Non-radiolabeled PK analysis

The PK properties of the four mAbs – mAb 1, mAb 2, mAb 3, and mAb 4—were evaluated in SCID-beige mice following a single IV bolus dose of 1 mg/kg administered through the tail vein. All four antibodies are human IgG1 isotypes with highly similar molecular weights and isoelectric points (pI), indicating their inherent structural resemblance. As expected, given their similar physicochemical properties, mAbs 1–3 exhibited comparable PK profiles (Figure 2). However, mAb 4 displayed a distinct PK profile (Figure 2), which we attribute to its lower levels of sialic acid on its Fab glycans (data not shown). The molecular properties of the four mAbs, including molecular weight and isoelectric point, are summarized in Figure 2. These mAbs are antiviral agents, designed to bind specific viral surface antigen targets that are not present in the SCID-beige mice used in this PK study. Therefore, this study primarily assesses the inherent PK and biodistribution properties of each mAb in the absence of target-binding interactions.

Figure 2.

Pharmacokinetic profiles and characteristics of antiviral mAbs in SCID-beige mice. Serum PK profiles of mAb 1, mAb 2, mAb 3, and mAb 4 over 7 days, measured by MSD-ECL. Each time point represents the mean ± SD of n = 3. Summary of molecular properties and PK parameters of the four mAbs is provided in the inset table.

The first molecule, mAb 1, targets the fusion glycoprotein of RSV. It exhibits a half-life of 8.79 ± 2.8 days and a clearance rate of 5.95 ± 0.9 mL/day/kg (Figure 2). The second molecule, mAb 2, targets HBV surface antigen. The half-life of mAb 2 was 8.32 ± 3.9 days, with a clearance rate of 6.28 ± 2.7 mL/day/kg. Next, mAb 3, which targets glycoprotein B of HSV, displayed a shorter half-life of 5.76 ± 1.4 days, and a higher clearance rate (11.20 ± 2.0 mL/day/kg) compared to mAbs 1 and 2. The most distinctive PK profile was observed with mAb 4, which targets HIV envelope protein gp120. It exhibited a notably short half-life of 1.74 ± 0.3 days and a high clearance rate of 61.80 ± 9.3 mL/day/kg from circulation (Figure 2). To summarize, mAbs 1 and 2 showed prolonged circulation with lower clearance rates, while mAb 3 demonstrated slightly faster clearance, and mAb 4 exhibited rapid clearance for an IgG with no host target.

Radiolabeled PK, biodistribution analysis

The biodistribution of the radiolabeled mAbs (mAbs 1–4) was evaluated in SCID-beige mice following a single IV dose of 1 mg/kg. Figure 3 illustrates the concentration profiles (%ID/mL in plasma and %ID/g in tissues). To understand what typical values would be for mAbs with desirable PK, the average %ID/mL or %ID/g of mAb for mAbs 1–3 was used and compared to the average %ID/mL or %ID/g of mAb 4 across multiple tissues from 0.25 h through 168 h. The shaded bands around the mean center line represent the range of values observed for each group (Figure 3).

Figure 3.

Time-dependent tissue distribution of antiviral monoclonal antibodies in SCID-beige mice. The biodistribution of mAbs 1–4 was evaluated in plasma and tissues at multiple time points post-administration. Data are presented as percentage of injected dose per mL of plasma (%ID/mL) or per gram of tissue (%ID/g). Mean values are plotted as solid lines, with shaded bands representing the range: blue band for mAbs 1–3 (n = 6/time point for mAb 1; n = 4/time point for both mAbs 2 and 3), and orange band for mAb 4 (n = 6/time point). An alternative visualization of the time-dependent tissue distribution on a linear scale is shown in Figure S2.

Immediately following administration, mAbs 1–3 exhibited the highest concentration in plasma, with an average of 89.7%ID/mL at 0.25 h (S.T1). A rapid decline in plasma concentration was observed over the initial 24 h, decreasing to 45.2%ID/mL at 6 h and 35.8%ID/mL at 24 h, indicative of distribution from the central compartment into peripheral tissues. Beyond 24 h, the rate of decline in plasma concentration slowed, characteristic of a sustained terminal elimination phase typical for antibodies with FcRn recycling. mAb 4 displayed notably different plasma kinetics. While plasma was also the initial site of highest concentration, the average %ID/g was notably lower at 0.25 h (59.4%ID/mL, S.T1) compared to mAbs 1–3. Furthermore, mAb 4 exhibited a rapid decline in plasma concentration, decreasing to 24.4%ID/g at 6 h, 1.57%ID/g at 72 h, and reaching 0.23%ID/g by 168 h. This accelerated clearance results in a much shorter circulatory half-life and considerably less systemic exposure compared to mAbs 1–3.

Organs with high perfusion or significant reticuloendothelial system (RES) activity demonstrated substantial early uptake and prolonged retention of mAbs 1–3. The spleen showed consistently high concentrations, peaking at 59.5%ID/g at 6 h and remaining elevated at 31.6%ID/g at 24 h. Similarly, bone marrow exhibited notable uptake, reaching 26.0%ID/g at 6 h and 21.0%ID/g at 24 h. The liver also displayed relatively high initial concentrations (18.1%ID/g at 6 h, 12.2%ID/g at 24 h). Kidneys showed considerable uptake, with a peak of 15.7%ID/g at 6 h. Lungs and heart, both highly vascularized, demonstrated early distribution, with values of 14.0%ID/g and 12.4%ID/g at 6 h, respectively. mAb 4 also showed high initial uptake in these organs, but demonstrated extremely rapid clearance from these tissues. For instance, liver concentration peaked at a 39.24%ID/g at 0.25 h (significantly higher than mAbs 1–3) but rapidly decreased to 6.01%ID/g at 6 h and 1.62%ID/g at 24 h, becoming almost negligible by 168 h (0.08%ID/g). The spleen followed a similar pattern, with high initial uptake (22.4%ID/g at 0.25 h) but rapid clearance to 6.76%ID/g at 6 h and 0.56%ID/g at 72 h. Kidney uptake was also significant initially (19.9%ID/g at 0.25 h) but declined sharply.

A group of tissues, including the bladder, small intestine, large intestine, and bone, exhibited moderate but sustained antibody concentrations for mAbs 1–3 throughout the study. Although initial uptake was not as high as in RES-rich organs, these tissues maintained detectable and moderate antibody levels even at later time points. For instance, the small intestine showed concentrations ranging from 2.96%ID/g at 6 h to 1.81%ID/g at 120 h. Bone maintained relatively stable concentrations, exemplified by 2.07%ID/g at 6 h and 1.04%ID/g at 168 h. For mAb 4, distribution into these tissues was generally much lower and clearance was significantly faster. While some initial uptake was observed (e.g., small intestine 1.77%ID/g at 0.25 h), concentrations rapidly dropped to very low levels by 72 or 120 h (e.g., small intestine 0.17%ID/g at 72 h, 0.06%ID/g at 120 h). This indicates poor and transient distribution to these peripheral compartments for mAb 4.

Consistent across antibodies, brain tissue exhibited the lowest %ID/g across all time points. For mAbs 1–3, brain concentrations ranged from 0.92%ID/g at 0.25 h to 0.27%ID/g at 168 h. Similarly, mAb 4 also showed negligible brain penetration, with concentrations of 0.82%ID/g at 0.25 h and an extremely low 0.01%ID/g by 168 h.

Across most tissues, concentrations for mAbs 1–3 peaked within the first 6 to 24 h, followed by a gradual decline over the subsequent time points, reflecting systemic clearance mechanisms. Tissues such as the spleen, liver, and bone marrow demonstrated relatively slower clearance rates, indicating prolonged retention. In sharp contrast, mAb 4 exhibited a rapid overall clearance from both plasma and nearly all tissues. Detailed biodistribution plots for each mAb across all time points is shown (Fig. S2). Analysis of the data revealed distinct patterns of antibody distribution and clearance, highlighting the differences between a presumably desirable PK profile (mAbs 1–3) and one exhibiting undesirable characteristics (mAb 4) (Figure 3).

Bulk total and interstitial concentrations of mAbs 1, 2, and 3 in tissues

To investigate tissue partitioning, we evaluated total and interstitial concentrations across 11 tissues at a quasi-steady-state of 168 h (Figure 4), comparing them to the average plasma concentration of mAbs 1, 2 and 3. Given its aberrant PK behavior, mAb 4 was excluded from the main visualization of total and interstitial tissue concentrations. Total tissue concentrations were highest in spleen (1.62 µg/mL) and bone marrow (0.91 µg/mL), and lowest in brain (0.04 µg/mL). Interstitial concentrations were generally lower than plasma levels but approached or exceeded plasma concentrations (4.26 µg/mL) in bone marrow (8.03 µg/mL), liver (4.32 µg/mL), and lung (3.70 µg/mL), indicating efficient interstitial penetration in these organs. Bone (1.19 µg/mL), large intestine (0.86 µg/mL), and small intestine (1.35 µg/mL) showed lower interstitial concentrations relative to plasma. Due to the inability to accurately determine the interstitial volumes for the brain, bladder, kidney, and spleen (as described earlier), interstitial concentrations for these organs were not derived. Detailed values for both bulk total tissue and interstitial concentration for all time points and tissues are reported in the supplemental material and shown in Fig S3.

Figure 4.

Bulk total tissue and interstitial concentration of mAb 1, 2, and 3 at 168 h. Concentration of mAb 1, 2, and 3 in total tissue and interstitial fluid across various tissues at 168 h post-administration. The blue dashed line represents the average plasma concentration (4.26 µg/mL) at the same time point. Data are presented on a logarithmic scale. Error bars represent the standard deviation (SD) of the measurements. Aggregate data from 14 mice are shown: n = 6 for mAb 1, n = 4 for mAb 2, and n = 4 for mAb 3.

Fold change analysis, calculated as the ratio of interstitial to total tissue concentrations, provided critical insights into tissue-specific mAb partitioning. This approach complements absolute concentration data by quantifying the efficiency of interstitial penetration, highlighting the effeciancy of extravasation and distribution within different tissue microenvironments. Our findings showed varying degrees of interstitial enrichment: heart (3.7-fold), lung (4.0-fold), large intestine (5.6-fold), small intestine (5.1-fold), bone (7.0-fold), bone marrow (8.8-fold), and liver (10.1-fold). These differences reflect biological variations in interstitial accessibility, suggesting a gradient of extravasation and distribution efficiency across tissues.

To assess the reliability of traditional bioanalytical methods for measuring total tissue concentrations, we compared LBA results with those obtained from radiolabeled biodistribution studies. This comparison was intended to evaluate whether LBA, commonly used in routine PK studies, can accurately reflect tissue exposure, and to highlight the advantages of radiolabeling in capturing total antibody-derived material. We used mAb 1 in a routine PK study and collected plasma, heart, liver, spleen, lung, brain, and large and small intestine to compare the results from the radiolabeled biodistribution study. Quantification of tissue and plasma from the routine PK study was carried out with the same method described previously for the non-radiolabeled PK analysis. Similar values for plasma PK were observed between the studies as noted earlier. When comparing total tissue concentrations obtained by LBA to those from the radiolabeled (I-125) biodistribution study, LBA measurements were consistently lower than the corresponding radiolabeled measurements across all tissues examined (Fig. S4). Our data underscore the importance of exercising caution when assessing total tissue concentrations via LBA, suggesting careful interpretation of such measurements.

In summary, mAbs 1, 2, and 3 exhibit heterogeneous tissue distribution at the quasi-steady-state of 168 h. The near-equivalent interstitial and plasma concentrations in bone marrow, liver, and lung highlight efficient partitioning into these compartments. Notably, the liver demonstrated a high interstitial to total tissue concentration ratio, directly addressing our aim to characterize partitioning in this key organ of interest.

Antibody biodistribution coefficients and partitioning ratios

We investigated the biodistribution of the four distinct monoclonal antibodies across 11 tissues in SCID-beige mice over 168 h, focusing on the dynamic relationship between plasma and tissue concentrations in the absence of specific viral targets. Initial examination of the time-dependent concentration profiles (Fig. S5) revealed both linear and non-linear relationships, suggesting a dynamic distribution phase preceding potential equilibrium in some tissues (heart, lung, bladder, bone, large and small intestine), while others exhibited a more linear correlation (bone marrow, brain, kidney, liver, spleen).

Recognizing that simple linear correlations did not fully capture the dynamic distribution observed in all tissues, we used an exponential limited-growth equation (Eq. 8) to quantitatively characterize these dynamic and steady-state aspects of biodistribution. We analyzed mAb 123 (a joint analysis of mAbs 1, 2, and 3, representing antibodies with normal and desirable PK) and mAb 4 to model the steady-state ABC and rate constant (K) values across the different tissues. (Table 2). As illustrated in Figure S6, the fitted curves accurately describe the experimental ABC data over time, where the ABC values converge toward a stable level, indicating that eventually an equilibrium and a fixed biodistribution coefficient within each tissue is reached. Additionally, Eq. 9 were used to model the steady-state PR and rate constant (K’) for mAb 123 and mAb 4 (Table 2).

Table 2.

Antibody Biodistribution coefficient (ABC) and Partitioning Ratios (PR) at steady-state and their equilibrium rate constants (t_1/2eq, t’_1/2eq) in various tissues.

Tissue	mAb	%ABCss Mean (Q195)	%PRss Mean (Q195)	t 1/2eq Mean (Q195)	t’ 1/2eq Mean (Q195)
Heart	123	15 (14–16)	58 (54–61)	0.17 (0.14–0.12)	0.25 (0.19–0.32)
	4	21 (20–22)	89 (83–95)	0.32 (0.27–0.32)	0.50 (0.43–0.63)
Lungs	123	21 (20–22)	83 (78–89)	0.16 (0.14–0.18)	0.29 (0.24–0.36)
	4	33 (30–35)	163 (149–180)	0.36 (0.32–0.41)	3.47 (2.31–6.93)
Lg Intestine	123	4 (4–4)	23 (21–25)	0.58 (0.53–0.63)	1.39 (0.87–2.31)
	4	15 (12–20)	94 (74–120)	0.77 (0.63–1.16)	0.77 (0.63–1.16)
Sm Intestine	123	7 (6–7)	35 (33–38)	0.35 (0.30–0.41)	0.53 (0.46–0.63)
	4	15 (14–16)	69 (64–76)	0.77 (0.63–0.99)	0.99 (0.77–1.39)
Bone	123	4 (4–4)	28 (27–30)	0.08 (0.08–0.09)	0.22 (0.17–0.32)
	4	3.5 (2.7–4.6)	63 (57–68)	0.58 (0.50–0.69)	0.77 (0.63–1.16)
Bone Marrow	123	24 (23–26)	258 (223–294)	0.02 (0.02–0. 02)	0.02 (0.01–0.22)
	4	18 (5.1–56)	159 (151–169)	0.11 (0.09–0.14)	0.12 (0.10–0.14)
Liver	123	11 (11–11)	118 (109–128)	0.02 (0.02–0. 02)	0.02 (0.01–0.11)
	4	66 (61–72)	890 (818–970)	0.19 (0.14–0.27)	0.20 (0.15–0.28)
Brain	123	1 (1–1)	NA	0.04 (0.03–0.06)	NA
	4	2 (2–2)	NA	0.26 (0.21–0.33)	NA
Kidney	123	16 (16–16)	NA	0.01 (0.01–0.01)	NA
	4	27 (25–29)	NA	0.18 (0.16–0.21)	NA
Spleen	123	42 (39–45)	NA	0.02 (0.01–0.08)	NA
	4	78 (72–85)	NA	0.16 (0.13–0.20)	NA
Bladder	123	18 (17–20)	NA	6.93 (3.47–6.93)	NA
	4	47 (39–57)	NA	0.99 (0.69–3.47)	NA

Steady-state ABC, representing the tissue-to-plasma concentration ratio at equilibrium was calculated for the 11 tissues (Figure 5, Table 2). For mAb 123, high ABC values were found in spleen, bone marrow, and lung, while brain and bone showed the lowest. mAb 4 exhibited different distribution coefficients, with the highest ABC in spleen and bladder, and low values in brain and bone. Notably, mAb 4 also demonstrated a substantially higher ABC in the liver compared to mAb 123 (Figure 5). This observation is consistent with mAb 4‘s potential as an ASGPR substrate due to its unique glycan profile, leading to more rapid uptake and distribution within the liver.

Figure 5.

Steady-state antibody biodistribution (ABC) and partitioning ratios (PR) coefficients across tissues. Bar plot depicting the ABC and PR coefficients for mAb 123 and mAb 4 in different tissues (error bars represent the QI95). Steady-state values were derived from the longitudinal data calculated using Eq. 8 and 9. Aggregate data from 196 mice are shown for mAb 123: n = 6/time point for mAb 1, n = 4/time point for mAb 2, and n = 4/time point for mAb 3. mAb 4 represents aggregate data from 84 mice: n = 6/time point.

To provide a more accessible metric for the speed of tissue-plasma equilibration, the equilibrium rate constants (K) were converted into corresponding half-lives (t_1/2eq), in hours, using Equation 10. These equilibrium half-life values offer a clear and quantifiable measure of the time required to reach quasi-steady-state conditions in different tissues, simplifying the interpretation of the antibodies’ unique disposition.

The estimated equilibrium half-lives, reflecting the speed of tissue-plasma equilibrium, revealed distinct kinetic profiles for mAb 123 compared to mAb 4 (Table 2). Surprisingly, for both antibody groups (mAb 123 and mAb 4), most tissues achieved 50% of their steady-state tissue-to-plasma ratio in under an hour.

For mAb 123, very rapid equilibration was observed in the kidney (t_1/2eq = 0.01 hr), bone marrow (t_1/2eq = 0.02 hr), liver (t_1/2eq = 0.02 hr), spleen (t_1/2eq = 0.02 hr), brain (t_1/2eq = 0.04 hr), and bone (t_1/2eq = 0.08 hr). Fast equilibration was noted in the lung (t_1/2eq = 0.16 hr) and heart (t_1/2eq = 0.17 hr). Tissues exhibiting intermediate equilibration rates included the small intestine (t_1/2eq = 0.35 hr) and large intestine (t_1/2eq = 0.58 hr). The bladder was a notable exception, displaying a significantly slower equilibration half-life of 6.93 h for mAb 123.

mAb 4 generally displayed higher equilibrium half-life values compared to mAb 123, indicating slower overall equilibration in all tissues except the bladder. Similarly, for t’_1/2eq (the equilibrium of the interstitial to plasma concentration) mAb 4 showed slower rates in all tissues except the large intestine, where it exhibited a faster rate than mAb 123.

The steady-state PR, representing the ratio of interstitial to plasma concentration at equilibrium (calculated using Eq. 9), also revealed tissue-specific differences (Figure 5, Table 2). For mAb 123, the highest PR values were observed in bone marrow and liver, suggesting a greater partitioning into the interstitial space relative to plasma in these tissues. Lower PR values were noted in brain and bone. mAb 4 showed a different pattern, with the highest PR in bone marrow and lung. The PR for mAb 4 in the liver was higher than that of mAb 123.

In summary, the analysis of ABC and partitioning ratios revealed distinct tissue-specific equilibration rates (t_1/2eq, t’_1/2eq) and steady-state distribution characteristics (ABC and PR) for mAb 123 and mAb 4 across the 11 investigated tissues (Figure 5, Table 2). The equilibrium distribution, as reflected by both ABC and PR, varied significantly between the two antibody groups, with mAb 4 showing a particularly high ABC and PR in the liver compared to mAb 123. Ultimately, these results highlight that comprehending the specific structural attributes of an antibody is paramount for accurately leveraging ABC and PR values in predicting its tissue-specific PK behavior

Discussion

This study provides a comprehensive characterization of the in vivo biodistribution of four monoclonal antibodies in SCID-beige mice, integrating detailed measurements of tissue vascular and interstitial volumes with PK time-course and biodistribution data. Our findings offer novel insights into the dynamic and steady-state partitioning of IgG antibodies across a wide range of tissues, addressing limitations of previous studies that often relied on extrapolated or limited time course data.

A significant strength of this study was the unique combination of experimentally determined tissue vascular (Vv) and interstitial (Vi) volumes with a detailed in vivo biodistribution analysis within the same experimental system. This integrated approach allowed for the direct calculation of interstitial antibody concentrations using physiology data generated in parallel. The comprehensive characterization of Vv and Vi across 19 organs and tissues provides a valuable resource, addressing the common reliance on potentially inaccurate literature-derived parameters from different preclinical species. We observed some sex-specific differences in the vascular (heart) and interstitial volumes (heart, small intestine, spleen), underscoring the importance of considering sex as a biological variable. Technical limitations prevented accurate interstitial volume determination in the brain, bladder, kidney, and spleen.

The PK analysis revealed distinct clearance profiles, with mAbs 1, 2, and 3 exhibiting typical IgG behavior, while mAb 4 showed significantly faster clearance and high distribution to the liver, due to its unique Fab glycosylation (data not shown). The comparable PK profiles between radiolabeled and non-radiolabeled mAbs validated the use of tracers for biodistribution studies. The consistent observation of lower total tissue concentrations by LBA compared to radiolabeling highlights significant methodological differences. While radiolabeling offers a comprehensive measure of all antibody-derived material present in the tissue, LBA primarily quantifies only intact, structurally unimpaired mAb that is successfully recovered in the supernatant during tissue homogenization and processing. The LBA process for tissue homogenates faces challenges, as mechanical forces can induce antibody aggregation or lead to tight association with insoluble tissue components, resulting in incomplete recovery. This systemic underestimation by LBA suggests that, while it may not be ideal for precisely quantifying absolute tissue levels, LBA can still provide valuable insights for comparative analyses, particularly when assessing the relative distribution of different molecules in preclinical studies. For general purposes and to overcome the limitations of incomplete recovery, empirically derived ABC and PRs offer a more robust and practical approach to inferring total tissue and interstitial concentrations from plasma data. These ratios reflect steady-state distribution characteristics and are less susceptible to the assay-specific recovery biases inherent in direct LBA measurements, thereby providing a more reliable framework for predicting antibody exposure and informing modeling and translational applications.

Time-course biodistribution data highlighted tissue-specific differences in antibody exposure. mAbs 1–3 exhibited sustained systemic exposure and desirable tissue retention, consistent with effective FcRn recycling. Conversely, mAb 4 demonstrated significantly accelerated clearance from plasma and tissues, particularly in RES-rich organs, indicating extensive nonspecific uptake and rapid catabolism, hallmarks of undesirable PK properties. Both antibody groups showed negligible brain penetration.

Our analysis of ABC over a time course, estimating t_1/2eq , t’_1/2eq , ABC, and PR of mAb123 offers a more detailed understanding of tissue-specific biodistribution dynamics than studies relying on limited steady-state data. A significant and novel observation from this work is that the total tissue-to-plasma (ABC) and interstitial-to-plasma (PR) ratios reached their half-maximal equilibration (t_1/2eq) much quicker than previously anticipated. Furthermore, the time to reach maximum tissue concentration (Tmax) also occurred more rapidly than expected, often as soon as 6 h for most tissues. Both findings are new observations that advance our understanding of antibody distribution kinetics.

Comparison of mAb 123‘s ABC values with previously reported data reveals varied agreement across tissues. Liver demonstrated a similar distribution ratio (−9%) to previously published data,⁸ suggesting consistent partitioning in this organ. In contrast, a range of tissues exhibited slight differences, including kidney (+17%), small intestine (+34%), lung (+41%), heart (+47%), and notably lower distribution ratios in bone (−45%) and large intestine (−20%). Most strikingly, different values were observed in the brain (+185%) and spleen (+228%), indicating substantially higher tissue-to-plasma ratios for mAb 123 in these tissues compared to the reference values.⁸ These discrepancies underscore the importance of experimental methodologies and specific antibody characteristics when interpreting and comparing tissue distribution ratios across studies

Variations in the equilibrium rate constants highlight the role of tissue barriers and transport in equilibration rates, with highly perfused tissues such as kidney equilibrating faster than bone and intestines. The observed differences in equilibrium rates between mAb 123 and mAb 4 can be attributed to mAb 4‘s abnormal PK profile, particularly its putative role as an ASGPR substrate and its high distribution to specific tissues, which collectively prolong the time required to reach equilibrium. The steady-state ABC revealed distinct equilibrium partitioning patterns, with high ABC for mAbs 1–3 in lymphoid tissues (spleen, bone marrow) and highly vascularized organs (lung) indicating extensive and efficient distribution into these tissues. The markedly higher ABC of mAb 4 in the liver further emphasizes its unique disposition. These findings are valuable for improving PK modeling and the translational relevance of preclinical simulations.

Building on this, our analysis of interstitial concentrations at the 168-h quasi-steady-state for mAbs 1–3 demonstrated efficient partitioning into compartments like bone marrow, liver, and lung, where interstitial concentrations generally approached or even exceeded plasma concentrations. This contrasted with lower interstitial concentrations relative to plasma in tissues such as bone and intestines, suggesting varying degrees of extravasation. Interstitial concentrations could not be calculated for the bladder, brain, kidney, and spleen due to limitations in determining their interstitial volumes. While FcRn-mediated recycling and species-specific differences in FcRn affinity may influence systemic clearance and intracellular retention, our study focuses on extracellular and interstitial antibody distribution, where FcRn interactions are less central. Accordingly, the ABC and PR values reported here are intended to describe inherent tissue partitioning.

Furthermore, the fold change analysis of interstitial to total tissue concentrations for mAbs 1–3 revealed distinct groupings consistent with known capillary permeability differences. Lower interstitial enrichment occurred in tissues with continuous endothelium (heart, lung), intermediate enrichment in tissues with more permeable or mixed vasculature (intestines, bone), and higher enrichment in tissues with sinusoidal capillaries (bone marrow, liver). This strongly suggests that capillary permeability plays a significant role in antibody partitioning, particularly in the absence of target binding.

Beyond modeling, our data can inform in vitro assay designs by characterizing tissue-specific equilibrium times and providing ABC values for physiologically relevant concentration selection. The observed differences in K values across tissues emphasize the importance of considering tissue-specific barriers and transport mechanisms in predicting antibody distribution kinetics.

By integrating a rigorous physiological characterization of tissue volumes with dynamic and quasi-steady-state antibody measurements, this study represents one of the most comprehensive efforts to generate and characterize ABC and PR values across multiple healthy tissues. One of the persistent challenges in translational pharmacology is bridging the gap between preclinical PK characterization and clinical outcomes. For mAbs, which rely on target engagement within specific tissues, systemic exposure alone may not reflect therapeutic potential. The biodistribution data presented in this study provide insight into tissue-specific concentrations that are often overlooked in standard PK assessments. These data can inform more accurate predictions of clinical efficacy, particularly when the site of action is not well-represented by plasma levels. The robustly derived coefficients offer a valuable baseline for interpreting tissue distribution in scenarios where plasma PK closely resembles the normal IgG profiles presented herein.

Importantly, the ABC and PR values generated in this study can serve as refined inputs to PBPK and modeling frameworks, improving the ability to estimate tissue-specific and interstitial concentrations. Rather than replacing these approaches, the data presented here enhance their physiological relevance and reduce reliance on oversimplified or static literature-derived parameters. This improved granularity supports more accurate extrapolation efforts and strengthens the translational bridge from preclinical biodistribution to human PK prediction. Nonetheless, the direct applicability of these coefficients to conditions involving target-mediated drug disposition or altered physiological states – such as inflammation, fibrosis, or tumor pathophysiology – may be limited. New coefficients would need to be specifically derived for such altered systems. This work provides a foundation for enabling more precise PK modeling and a deeper mechanistic understanding of antibody distribution in complex biological systems. Our findings, particularly the demonstration that interstitial concentrations in key organs like the liver can mirror or even exceed plasma levels depending on mAb properties, underscore the importance of evaluating sub-compartmental distribution. This highlights how total tissue concentrations alone can underestimate the therapeutically relevant exposure at the site of action.

This study provides a comprehensive dataset and analysis of antibody biodistribution, offering valuable insights into the dynamic and steady-state partitioning of IgG antibodies across multiple tissues. Our findings contribute significantly to the understanding of tissue-specific antibody PK, providing a foundation for improved translational modeling, in vitro assay design, and a more nuanced appreciation of the factors governing antibody distribution in vivo. The detailed characterization of vascular and interstitial volumes, coupled with the time-course analysis of ABC, addresses limitations in the existing literature and provides a more precise assessment of antibody biodistribution at the tissue and sub-compartmental levels.

Funding Statement

The author(s) reported there is no funding associated with the work featured in this article.

Disclosure statement

All authors except Leslie Khawli are current or former employees of Gilead Sciences. Leslie Khawli is a consultant for Gilead Sciences.

Significance statement

This study provides a comprehensive dataset of experimentally derived vascular and interstitial volumes, radiolabeled antibody PK, and biodistribution across multiple tissues. By calculating total and interstitial tissue concentrations over time, along with steady-state ABC and PR values, this work significantly advances our understanding of interstitial antibody PK. This detailed dataset provides a more accurate foundation for preclinical PK modeling, ultimately improving the prediction of in vivo efficacy.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/19420862.2025.2587580